feat: connect wandb

Referenced Issue: #2

model/font_classifier/train_font.py

@@ -11,6 +11,7 @@
 import matplotlib.pyplot as plt
 import numpy as np
 import torch
+import wandb
 
 from scheduler import scheduler_module
 from torch.utils.data import DataLoader
@@ -133,25 +134,6 @@ def train(data_dir, model_dir, args):
     # --scheduler
     if args.scheduler != "None":
         scheduler = scheduler_module.get_scheduler(scheduler_module,args.scheduler, optimizer)
-
-    # -- logging
-    logger = SummaryWriter(log_dir=save_dir)
-    with open(os.path.join(save_dir, 'config.json'), 'w', encoding='utf-8') as f:
-        json.dump(vars(args), f, ensure_ascii=False, indent=4)
-
-    layout = {
-        "Train_Val": {
-            "accuracy": ["Multiline", ["Train/accuracy_epcoh", "Val/accuracy"]], 
-            "f1_score": ["Multiline", ["Val/f1_score"]], 
-            "loss": ["Multiline", ['Train/loss_epoch', 'Val/loss']]
-        },
-    }
-
-    logger.add_custom_scalars(layout)
-
-    best_val_acc = 0
-    best_val_loss = np.inf
-    best_f1_score = 0
 
     for epoch in tqdm(range(args.epochs)):
 
@@ -186,15 +168,21 @@ def train(data_dir, model_dir, args):
                     f"training loss {train_loss:4.4} || training accuracy {train_acc:4.2%} || lr {current_lr}"
                 )
                 # logs
-                # logger.add_scalar("Train/loss", train_loss, epoch * len(train_loader) + idx)
-                # logger.add_scalar("Train/accuracy", train_acc, epoch * len(train_loader) + idx)
-                # logger.add_scalar("lr", current_lr, epoch)
+                wandb.log({
+                    "Train/loss": train_loss,
+                    "Train/accuracy": train_acc,
+                    "step" : epoch * len(train_loader) + idx + 1
+                    })
 
 
         loss_value_sum = loss_value / args.log_interval
         train_acc_sum =  matches / args.batch_size / len(train_loader)
-        logger.add_scalar("Train/loss_epoch", loss_value_sum, epoch)
-        logger.add_scalar("Train/accuracy_epcoh", train_acc_sum, epoch)
+        wandb.log({
+            "Train/loss_epoch": loss_value_sum,
+            "Train/accuracy_epcoh": train_acc_sum,
+            "lr": current_lr,
+            "epoch": epoch
+            })
 
 
         # val loop
@@ -231,9 +219,11 @@ def train(data_dir, model_dir, args):
 
             print(f"[Val] acc : {val_acc:4.2%}, loss: {val_loss:4.2}, f1: {f1_score:4.4} ")
 
-            logger.add_scalar("Val/loss", val_loss, epoch)
-            logger.add_scalar("Val/accuracy", val_acc, epoch)
-            logger.add_scalar("Val/f1_score", f1_score, epoch)
+            wandb.log({
+                "Val/loss": val_loss,
+                "Val/accuracy": val_acc,
+                "Val/f1_score": f1_score
+            })
 
         if (epoch + 1) % args.save_interval == 0:
             # model save
@@ -266,13 +256,22 @@ def train(data_dir, model_dir, args):
     parser.add_argument('--save_interval', type=int, default=5, help='how many epochs to wait before save pth')
 
     # Container environment
-    parser.add_argument('--data_dir', type=str, default='/opt/level3_productserving-level3-cv-11/data/words/ko/images') 
+    parser.add_argument('--data_dir', type=str, default='/opt/level3_productserving-level3-cv-11/data/words/ko/typical_images') 
     parser.add_argument('--model_dir', type=str, default = './experiment/classifier')
-
+
+    # wandb
+    parser.add_argument('--tags', default= None, nargs='+',type=str, help = "프로젝트 태그 할당")
     args = parser.parse_args()
     print(args)
 
     data_dir = args.data_dir
     model_dir = args.model_dir
-
+
+    wandb.init(entity = "miho",
+               project = "Final-Project",
+               sync_tensorboard=True,
+               name = args.name,
+               tags = args.tags)
+    wandb.config.update(args)
+
     train(data_dir, model_dir, args)